基于頻域系統(tǒng)辨識(shí)的石油K型井架結(jié)構(gòu)損傷理論研究.pdf

1、大慶石油學(xué)院碩士學(xué)位論文基于頻域系統(tǒng)辨識(shí)的石油K型井架結(jié)構(gòu)損傷理論研究姓名：王樹慶申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：安全技術(shù)及工程指導(dǎo)教師：鄒龍慶20070318ResearchinDamageDetectionTheoryofKTypeDerrickStructureBasedollFrequencyDomainSystemIdentificationrigderrickisoncofthemostimportantequipmentinhoi

2、stsystemPeoplecminstallcrownblockandhangtacklehoistinghookandspecialtoolsonitSomederricksinusinghavebeenusedforalongtimeintheopenmanypartsofthederrickhavedistortionandcorruptionofdifferentcxtenlendenlargedscrewThuswemust

3、detectende刪uakthedamageandfullyrealizctheactualstatusofthederrick醴luctufein璐in參Ifwecanfindthedamageimmediatelydiagnosethelocationandtheextentofthedamage，constitutethecorrectservicingstra忙gy’gattlmetheabilityofbearingthew

4、eightoftoaditisgoodtopostponetheBaturallifeofderrickandtoprovidetheproofofusingpresentderrickeconom／callyFurthermore，itisveryimportanttoavoiddisasterandtoprotectthepeopleslivesandestateThe咖dyondamageidentificationofderri

5、ckandmakingprogressin酬uatiI喀theabilityofloadendsafetyhasimportantrealisticvalueandthcofedcssignificanceWustlyonthebasisofcollectionandanalysisofmuchliterature，thisarticleanalysesthetechnologyofsUucturaldamageidentificati

6、onofdomesticendforeignresearchpresentsituationandintroducedthebasictheoryofmodalexperimentandthebasictheoryendmethodofthederricksmodalparameterbasedonfrequencydomainsystemidentificationSecondlythedamageidentificationmeth

7、odbasedonBPneuralnetworksisdiscussedandamly∞tAllofthesesuchasthebasiceonsoptofartificialneuralnetworksthecharacteristicthemodelitsalgorithmarcintroducedinthisprocessAtthesamedI∞ThisarticleelaboratesBPneuralnetworksBPclas

8、sicalgorithm’theLavenderMarquardtal鯽thmontheconditionofoptimizationandBPI^euialnetworksthatreafizedintheM怕TLABind酏lilsIngeneralaimingattheproblemonthedamagelocationendextentofJJl60／41Ktypederrickthemethodofdamageidentifi

9、cationbasedonartificial∞llralnetworksisintroducedVibrationmodalanalysistheoryisintegratedwithBPneuralnetworkssoastodetectderricksdamagewiththehelpofANSYSandMATLAB，atthesamctime，ithasbeensuccessfulinde把c吐ngdamagedposition

10、andextentlikelyoccurringthemethodbasedonvibrationmodalanalysistheoryandBPneuralnetworksissatisfiedinthispaperAlongwiththeimprovementofprecisionofthemodalanalyticapparatusandthetechnologyofidentifyingparametertheresultsis